Harmonizing data dictionaries across all health data collection, sharing, and reporting initiatives

Health Information Technology could improve the quality of healthcare and reduce its costs, according to a recent report by the President's Council of Advisors on Science and Technology (PCAST). One of PCAST conclusions states that to achieve these objectives, the Federal Government takes the role of facilitator of the nationwide adoption of a universal exchange language for healthcare information and a digital infrastructure for locating patient records while strictly ensuring patient privacy. A universal exchange language could also facilita data sharing between public health and clinical care when data standards are pervasive. These efforts imply the harmonization of data standards for health transactions at the point of care. To date we have relied on successful national efforts to assess performance and outcomes from national databases. Several Federal Government and national organizations sponsor data collection, sharing, and reporting without harmonization of data standards. This study will present results from an analysis of data dictionary elements currently used by NCHS' Health Care Data Surveys; AHRQ's HCUP and MEPS; CMS' MDS, HCAHPS, UHDDS, and other widely used health care databases. Our findings show that a minimum of five different formats exists to capture sex/gender variables. The disharmonious formats are more evident in other important demographic variable and value labels. To achieve HIT interoperability and meaningful use we need to extend data standards beyond EHRs to all health data collection, sharing, and reporting initiatives.